76 Claude Fuller 



lower elevations, (Pretoria 4,400, Pienaars River 3,500 ft.)- 

 Those again from the lower parts of the Transvaal (Tzaneen 

 2,300 ft.) are shorter still. The imagos from LourenQo Marques 

 fall into the last mentioned series. TTie impression one gathers 

 is that wings tend to be longer at higher than lower altitudes but 

 this is rather set at naught by the long winged imagos from 

 Inchanga. 2,000 ft. 



Tentatively, I propose to recognise three iotms cf M. natalensis. 



1. M.natalensis s.str. characterized by its tendency to vary 



and by the frequency with which thick headed soldiers 

 with angular mesonota are present. Normally a builder 

 of conical mounds, 3 to 4 ft, high. Span, 68 to 81 mm. 



2. M.natalensis form transvaalensis characterized by the higher 



wing measurements of the imago, the stability of the 

 soldiers and the general absence of such with angular 

 mesonota. Normally a builder of low rounded mounds. 

 Span 74 to 80 mm. 



3. M. natalensis form intermedins, not distinguishable from 



form transvaalensis except that the Imagos have shorter 

 wings and insignificantly shorter heads with somewhat 

 more prominent ocelli. This form regularly builds high 

 to very high conical mounds (6 ft. to 12 ft.) and in this 

 stands, in a broad biological sense, between natalensis 

 and l)ellicosus.. Span, 72 to 74 mm. 



Localities: Cape Province: (Bolo, Kei River; Kimberley). 

 Orange Free State: (Reitkuil; Vierfontein) . Transvaal; 

 (Barberton, Bethal, Duivel's Kloof, De Wildt, Johannesburg, 

 Krugersdorp, Lichtenburg, Letsitele, Leydsdorp, Louis Trichardt, 

 Mokeetse, Nylstroom, Naboomspruit, Pietpotgietersrust, Pretoria, 

 Pietersburg, Pienaar's River, Rustenburg, Tzaneen, Waterval 

 Onder, Woodbush, Witbank, Warmbaths). NatAL: (Bellair, 

 Durban, Dumisa, Estcourt, Hillcrest, Haviland Rail, Mount 

 Edgecombe, Muden, Ladysmith, New Hanover, Pietermaritz- 

 burg, Winterton, Winklespruit). ZuLULAND: (Lower Ukuzi, 

 Mseleni, Lake Sibayi, Lower Umfolosi, Mtunzini). 



Note. M. bellicosus and M. natalensis are undoubtedly very 

 closely allied species both in structure and biology. However 

 the two stand well apart both as regards imagos and soldiers and 

 the two are not readily confused. Here it may be mentioned that 

 within the boundaries of the Union of South Africa each species 

 is restricted to its own terrain and I have not so far found any 



